We're Hashicorp partners, however up until now we've kept our experience and deployments firmly in the cloud. We wondered if the Hashicorp tooling would help us in other situations, and found that it performed even better than we had hoped.
Matt Wallace (one of our Directors) decided to take the plunge over the Christmas and New Year break and apply the Hashicorp stack (plus a few other goodies) to his home management setup...
Up until very recently, all of the home management devices that I used were running on a multitude of Raspberry Pi's or similar devices. I had one pi for Home Assistant to run my home automation, another for NXFilter to keep my kids safe online, more still to manage my 3D printer and CNC Machine, and a couple connected to the network "just in case" for experimentation purposes.
I'd mounted them on a bit of 9mm Plywood, along with power supplies and a network switch, and it had been working fine, until I upgraded the networking kit to Ubiquiti Unifi.
All of a sudden, I had managed switches that would deal with 1Gbps traffic, and the number of devices connecting via WiFi was increasing too. The home management side of things needed an upgrade to match the new challenges.
I've been working in Systems Administration and automation for nearly 20 years so I knew that whatever I replaced the Raspberry Pi's with, it would need to be easy to maintain and even easier to add to.
As much as I enjoy configuring software and infrastructure during the working day, when I'm at home I just want things to work without intervention.
The idea that I could run a "mini cloud" and launch the various applications into it gathered pace and I found myself evaluating several options including Openstack (far too powerful for this particular use-case!), Kubernetes (probably complete overkill because I don't need the distributed clustering etc), Docker Swarm (I've been bitten hard by this in a production environment in the past, so didn't really want to experience those pains again!), before finally turning to the Hashicorp stack.
Hashicorp's stack comprises of Vault for managing secrets, Consul for managing DNS and service discovery, and Nomad for managing containers and other jobs. That solved the "private cloud" aspect so all I needed now was a way to map the various ports to HTTP(S) without thinking about it, and that's where Traefik came in to the equation.
Vault tightly controls access to secrets and encryption keys by authenticating against trusted sources of identity such as Active Directory, LDAP, Kubernetes, CloudFoundry, and cloud platforms. Vault enables fine grained authorization of which users and applications are permitted access to secrets and keys.
In the past, we've used it to secure all kinds of information from Cryptocurrency-based Escrow systems to one-time-passwords for major financial institutions. My plan was to replace any credentials that might need to be stored in a container or in Ansible with a call to the Vault API.
Consul is a service mesh solution providing a full featured control plane with service discovery, configuration, and segmentation functionality. Each of these features can be used individually as needed, or they can be used together to build a full service mesh.
We've used Consul extensively in the past for everything from scaling out logging clusters through to service discovery for clustered IoT solutions, so I knew it would easily cope with anything my home network required.
HashiCorp Nomad is a single binary that schedules applications and services on Linux, Windows, and Mac. It is an open source scheduler that uses a declarative job file for scheduling virtualized, containerized, and standalone applications.
I'd used Nomad previously when advising a FinTech organisation on their infrastructure and CI/CD platform, and I fell in love with the powerful configuration language almost immediately! (more on that later...)
A reverse proxy / load balancer that's easy, dynamic, automatic, fast, full-featured, open source, production proven, provides metrics, and integrates with every major cluster technology, Traefik is a relative newcomer to the world of reverse proxies.
We're a busy household, so I also wanted to know when there was an issue with the system. Naturally, as we're also DataDog partners, I installed the DataDog and configured the integrations for Consul, Vault, Docker, and Traefik so I could monitor and get alerts if something failed.
The architecture of the system based on the criteria was relatively simple. The "core" services would run on the physical host, with everything else running in containers.
This meant installing Consul, Docker, Nomad, Vault, and Traefik onto the server, and then writing Nomad jobs for HomeAssistant, OctoPi, and the other services to take advantage of the underlying "infrastructure". It also allows me to "extend" the platform in future by adding more physical servers and installing Docker/Nomad on them to join them into a cluster of container nodes.
All DNS Queries for the local network would be handled by Consul, with any records for which it was not authoritative being handed off to the global DNS network. Thankfully, with DNSMasq this was going to be easy to do!
Ansible has been my configuration management tool of choice for many years now, and as I had a clear idea of what I needed to install, I headed over to Ansible Galaxy to find roles that I could use to install the base components before starting to launch the containers...
In part two of this series, I'll talk more about how the various components are configured to interact with each other, and show just how easy it is to launch multiple containers with automatic DNS resolution using Nomad, Docker, Consul, and Traefik.
If you'd like to learn more about how our Hashicorp partnership can help you manage your datacentre, cloud platform, or even your house more effectively, just fill in the quick form below and we'll be in touch very shortly.